HOW ARE SHORTS INFORMED? SHORT SELLERS, NEWS, AND … · 2020-07-07 · Kenan-Flagler Business...

Electronic copy available at: http://ssrn.com/abstract=1535337

HOW ARE SHORTS INFORMED?

SHORT SELLERS, NEWS, AND INFORMATION PROCESSING*

Joseph E. Engelberg

Kenan-Flagler Business School, University of North Carolina

[email protected]

Adam V. Reed


[email protected]

Matthew C. Ringgenberg


[email protected]

FEBRUARY 22, 2010†

ABSTRACT

Combining a database of short sellers’ trading patterns with a database of news releases, we

show that short sellers’ trading advantage comes largely from their ability to analyze publicly

available information. Specifically, the prior finding that short sellers’ trades predict future

negative returns (e.g., Boehmer, Jones, and Zhang (2008) and Asquith, Pathak, and Ritter

(2005)) is more than twice as strong in the presence of news stories. Further, the most profitable

short sales do not appear to come from market makers, but from clients, and these client short

sales are particularly profitable in the presence of news. We also show that the ratio of short

sales to total volume is nearly constant around news periods, and when we do find differences

between the timing of short sellers’ trades and the overall market, we find that relative to other

types of trading there is a significant increase in short selling after news stories. Finally, short

sellers’ ability to predict returns appears to be concentrated in many of the news categories in

which short sellers trade relatively late; a finding consistent with the idea that short sellers’

advantage arises from their ability to process publicly available information.

*The authors thank Paul Tetlock for assistance with the Dow Jones news archive and we thank Dow Jones for

providing access to their news archive. We have benefited from comments from Greg Brown, Jennifer Conrad.

Wayne Ferson, Günter Strobl and Robert Whitelaw. We also thank seminar participants at the University of North

Carolina and the 2010 Utah Winter Finance Conference. This paper was previously titled “Buy on the Rumor,

[Short] Sell on the News: Short Sellers, News and Information Processing.” †Comments welcome. © 2010 Joseph E. Engelberg, Adam V. Reed, and Matthew C. Ringgenberg.

Electronic copy available at: http://ssrn.com/abstract=1535337

1

There is overwhelming evidence that short sellers are informed traders. In particular, a

number of empirical papers find that short selling predicts future returns (Asquith and Meulbroek

(1995), Senchack and Starks (1993), and Boehmer, Jones, and Zhang (2008)). Return

predictability, however, tells us little about how short sellers obtain an informational advantage

over other traders. In this paper we address this question by combining a database of public

news events with a database of all short sale trades, a unique combination that allows us to

comprehensively examine the relation between short selling and the release of public

information.

One aspect of the relation between short sales and news that has received a lot of

attention in the literature is timing. Short sellers have been shown to trade before public

information is released. For example, Karpoff and Lou (2009) show that short selling increases

before the initial public revelation of firms’ financial misrepresentation. Similarly, Christophe,

Ferri, and Angel (2004) find evidence of informed short selling in the five days before earnings

announcements. The financial crisis has also been linked to the timing of short sellers’ trades,

with the Securities and Exchange Commission suggesting that short sellers spread “false rumors”

in an effort to manipulate firms “uniquely vulnerable to panic.”1

To examine whether short sellers’ informational advantage is due to timing, we begin by

looking for evidence of abnormal short selling ahead of news events in the U.S. over the 2005 to

2007 period, a pattern that would be consistent with anticipation. We find no such pattern. In

fact, we find that the ratio of short sales to total volume is nearly constant around news events.

Further, when we do find differences between the timing of short sellers’ trades and the overall

1 “What the SEC Really Did on Short Selling,” by Chairman Christopher Cox, 24 July 2008, The Wall Street

Journal.

2

market, we observe that, relative to other types of trading, there is a significant increase in short

selling after the news event. This result indicates that, on average, short sellers trade on publicly

available information, that is, they do not uncover and trade on information before it becomes

public.

Given the finding that short sellers trade on publicly available information, we next

explore whether short sellers’ informational advantage is due to their superior ability to process

public information. Several papers find that abnormal short selling or high short interest

unconditionally predicts lower future returns (see, e.g., Asquith and Meulbroek (1995), Senchack

and Starks (1993), Boehmer, Jones, and Zhang (2008)). We find that abnormal short selling

does indeed lead to lower future returns, but that this effect is largely concentrated around news

events: short selling’s predicative effect on future returns is more than twice as strong in the

presence of news stories. Thus, a short seller’s most informative trades appear to be those in

response to newly released public news, which is consistent with short sellers being good

processors of information.

An alternative explanation for the above result may be that some buyers make systematic

mistakes around news events (Antweiler and Frank (2006)), and that these buyers’ mistakes are

reflected in market makers’ offsetting short sales. To determine whether short sellers’ trades are

due to superior information processing or to offsetting positions, we exploit a unique feature of

the short selling data, namely, exempt versus non-exempt trade marking, to distinguish market

makers from non-market makers (or clients). We find that clients’ trades are particularly well

informed, and that these trades are much more profitable in the presence of news events. In

contrast, market makers’ trades are not particularly well informed, and there is no differential

impact in the presence of news. Overall, we conclude that the most informed short sales are

3

from clients, and that these shorts are particularly well informed in the presence of recent news.

This evidence lends support to the view that short sellers’ information advantage is due to their

superior information processing ability.

In our next set of tests we identify which types of information are associated with short

sellers’ advantage. To do so, we use the news subject classification in the Dow Jones archive to

sort stories into various categories ranging from analyst comments to earnings announcements to

new debt issues. We find that short sellers’ most informative trades are concentrated in five

categories: Corporate Restructurings, Earnings, Earnings Projections, New Products &

Services, and Stock Ownership. Further, many of these categories correspond to the categories in

which short sellers’ trades are measurably later than other investors’ trades, which lends

additional support to the idea that short sellers’ advantage stems from superior ability to process

publicly available information rather than an ability to uncover information before it becomes

publicly available.

Finally, we examine the economic significance of traders’ ability to trade on news by

implementing a portfolio approach. Recognizing that the presence of news is likely correlated

with firm characteristics and that some categories of news may be more relevant for some firms

than for others, we conduct an experiment in which each firm’s response to a news event is

matched by a similar firm’s response on the same day. We find that across all news categories,

short sellers’ advantage in predicting returns is concentrated in firms with news.

The findings of this paper shed light on the broader debate about the informational effects

of news announcements. While several papers argue that public news announcements reduce

information asymmetry (e.g., Korajczyk, Lucas, and McDonald (1991), Kacperczyk and Seru

4

(2007)), another thread of the literature argues that public information events present profitable

trading opportunities for skilled information processors (e.g., Engelberg (2008), Demers and

Vega (2008)), in effect increasing the asymmetry of value-relevant trading signals. Tetlock

(2009) weighs in on this question by modeling liquidity shocks and news and finds empirical

evidence suggesting that public information plays a key role in informing a subset of investors.

By focusing specifically on short sellers, this paper has a unique ability to shed light on the trade-

level evidence in this debate for two reasons: first, a number of papers show that short sellers are

informed (e.g., Asquith and Meulbroek (1995), Diamond and Verrecchia (1987)); second, short

sellers’ trades are among the few classes of trades that are uniquely identified. Our finding that

short sellers’ trades are more than twice as profitable in the presence of recent news is strong

evidence in favor of the idea that news presents profitable trading opportunities for skilled

information processors.

The remainder of this paper proceeds as follows. Section I discusses related literature.

Section II describes the databases used in this study. Section III presents our analyses and

findings. Finally, Section IV concludes.

I. Related Literature

The ideas in this paper relate to three distinct branches of the existing literature. First,

this paper relates to an extensive literature on the behavior of short sellers relative to other

traders. Second, our paper contributes to a growing literature on how market participants

respond to public news. Finally, this paper sheds light on an emerging debate on whether news

increases or decreases information asymmetry. In this section, we first discuss prior papers that

5

connect news to short selling. We then provide an overview of the relevant literature in each of

these three branches.

Several extant papers look at short selling behavior in the context of a single type of

corporate news event. As such, these studies shed light on a subset of this paper’s sample of

news events. Karpoff and Lou (2009), for example, examine short sellers’ positions in firms that

are investigated for financial misconduct; they find that short sellers generally anticipate public

announcements of investigations. Focusing on short sellers’ trades around earnings

announcements, Christophe, Ferri, and Angel (2004) find that short sellers do not tend to trade

before earnings announcements. Similarly, Daske, Richardson, and Tuma (2005) look at short

selling around earnings announcements and management forecast announcements and find no

evidence that short sale transactions concentrate prior to bad news events. Nagel (2005) looks at

the cash flow news implied by a vector auto regression and finds an asymmetric effect on

returns, indicating that short sellers help incorporate news into prices when short selling is not

constrained. Finally, Edwards and Hanley (2008) examine short selling around IPOs, a

newsworthy corporate event, and find evidence that casts doubts on short sale constraints as an

explanation for IPO pricing anomalies.

In contrast to the above papers, which identify patterns in short selling around specific

corporate new events, the current paper aims to uncover patterns in short sellers’ trades around

all types of corporate news events. In doing so, we try to understand short sellers’ behavior more

generally.

A. Short Sellers’ Trading Patterns

6

Several papers compare the trades of short sellers to the trades of other market

participants. There are several dimensions over which trades can be compared. Much of the

recent literature focuses on the profitability of trades, which roughly speaking, can be measured

as the performance of a stock’s price after the short sale. One of the most widely cited results in

this vein of the literature is found in Asquith and Muelbroek (1995), who show high short

interest precedes negative future returns, consistent with informed trading. Similarly, Asquith,

Pathak, and Ritter (2005) show that when short selling is constrained and there are relatively

diverse opinions, in some cases abnormally high short interest precedes negative future returns.

Using transaction data at a higher frequency, Boehmer, Jones, and Zhang (2008) find that

heavily shorted stocks significantly underperform lightly shorted stocks, and Diether, Lee, and

Werner (2008) show that not only do prices follow short selling, but short selling also follows

prices, that is, short sellers tend to short after price run ups. These results further indicate that

short sellers may have an informational advantage.2

In sum, the prior work above establishes that the performance of short sellers’ trades

indicates that short sellers’ trades are informed. Our paper contributes to this literature by asking

how short sellers come to enjoy an informational advantage in the first place.

B. Public News

2 A closely related dimension of research is whether short sellers’ trades reveal information to other market

participants. In other words, are short sellers’ trades news worthy in and of themselves? Senchack and Starks

(1993) show that abnormally large short interest announcements have small but significant negative returns.

Similarly, Aitken, et al. (1998) show that short sales are followed by price declines within 15 minutes on the

Australian Stock Exchange.

7

While a large literature examines volume and return phenomena around specific news

events (e.g., earnings announcements, mergers, and dividend initiations and omissions), a more

recent literature considers such phenomena around any corporate news event. Categorizing all

Wall Street Journal stories, between 1973 and 2001, Antweiler and Frank (2006) find that return

responses vary widely across news categories, although they find evidence of overreaction

(return reversal) on average. Also using a database of all news events, Tetlock (2008) finds

evidence of even stronger return reversal following repeated news events consistent with the idea

that investors overreact to “stale” news stories. Furthermore, using comprehensive news

databases, several studies examine whether well-known asset pricing anomalies are related to

news. Chan (2003) considers the momentum anomaly among stocks with and without recent

news and finds evidence of price momentum only among news stocks. Similarly, Vega (2006)

finds more earnings momentum among stocks with high differences of opinion on news days.

More recently, researchers have asked whether the content of news stories contains

value-relevant information. Tetlock, Saar-Tsechansky, and Macskassy (2008) and Engelberg

(2008) show that, indeed, the qualitative content of the information contained in news stories can

predict both earnings surprises and short-term returns. These findings support the idea that there

is value-relevant or “soft” information in news stories that is not immediately impounded into

prices.

To summarize, this literature highlights the importance of looking at more than one news

category in assessing short sellers’ behavior, and shows that the information content of news

leaves room for traders with different abilities to process the information to arrive at different

conclusions about the value relevance of the news. Our work builds on these findings by

analyzing the universe of corporate news events in the U.S. over our sample period, and by

8

asking whether, in our sample, information processing ability plays a role in the performance of

short sellers’ trades.

C. Public News and Informed Trading

There are two views regarding the relation between public news, such as the articles

available in the Dow Jones archive, and the trading of skilled investors. Under the first view,

public information does not provide traders with an information advantage, that is, managers

who rely on public information (rather than generate private information) are low-skilled.

Consistent with this view, Kacperczyk and Seru (2007) estimate managers’ reliance on public

information (RPI) as the R-squared of a regression of percentage changes in fund managers’

portfolio holdings on changes in analysts’ past recommendations and find that fund managers

with low RPIs (low reliance on public information) perform better than fund managers with high

RPIs (high reliance on public information).

Under the alternative view, the public release of information presents trading

opportunities for skilled processors of information, that is, when news is released, traders with

superior information processing skills can convert this news into valuable information upon

which to trade. Earnings announcements, for example, are often accompanied by lengthy

documents and conference calls that are scrutinized by information processors. Those traders

who show exceptional skill in converting such data into value-relevant information are rewarded

with superior returns on event-driven trades. Evidence consistent with this view comes from

studies that attempt to look at the textual content of news and firm announcements. Specifically,

Tetlock et al. (2008), Engelberg (2008), Demers and Vega (2008), and Feldman et al. (2009) all

9

show that the content of corporate news predicts returns, which is consistent with the view that

information processing skills can generate superior returns.

Our paper sheds light on the above debate by finding additional evidence in support of

the second view, that is, by showing that trades occurring after the release of news stories can be

more profitable than trades in non-news periods.

II. Data

The data used in this study come from the intersection of two databases. The first

database contains information on short sales while the second contains news articles from the

Dow Jones News Service. Below we describe the two databases in turn.

A. Short Sales

Information on short sales transactions comes from the NYSE TAQ Regulation SHO

database. Regulation SHO was adopted by the SEC in June of 2004 to establish new rules

governing short sales in equity transactions and to evaluate the effectiveness of price test

restrictions on short sales. As one consequence of regulation SHO, transaction-level short sales

data were publicly disclosed for the period January 3, 2005 through July 6, 2007. The NYSE

TAQ regulation SHO database therefore contains data for all short transactions that were

reported on the NYSE during this period. Specifically, the database contains the stock ticker, the

date and time of the transaction, the number of shares traded, the execution price, and an

indicator that denotes whether the transaction was exempt from price test rules. One of the

10

reasons a short sale transaction could be classified as exempt is that it was made by market

makers engaged in bona fide market making activity. The exempt indicator has thus been used

to separate trading by market makers from trading by non-market makers (e.g., Evans et al.

(2009), Christope, Ferri, and Angel (2004), Boehmer, Jones, and Zhang (2008), Chakrabarty and

Shkilko (2008)).3 However, when regulation SHO was implemented, a group of randomly

selected stocks was selected to be part of a pilot study for which the exempt/non-exempt

classification was no longer required. We exclude these pilot firms when using the exempt

indicator variable in our analyses.4

For the purposes of our analysis, we aggregate the transaction data at the daily level, and

we use the TAQ master files to add CUSIPs to the database. We then use the CRSP Daily Stock

Event file to add PERMNOs to the database. Finally, we add returns, total volume, and shares

outstanding information that we obtain from CRSP.

B. Dow Jones Archive

To compile our sample of news events, we use the Dow Jones archive as in Tetlock

(2009). This archive contains all Dow Jones News Service stories and Wall Street Journal

stories over our 2005 to 2007 sample period.

The Dow Jones database also contains subject codes that identify the information content

of each news article; for example, there is a code to indicate that an article contains information

3 For example, NASD NTM 06-53 notes that “Rule 5100(c)(1) provides an exception to the bid test for short sales

by a market maker registered in the security in connection with bona fide market making activity.” 4 Details regarding the regulation SHO pilot study, including a list of firms involved, are available on the SEC

website: http://www.sec.gov/rules/other/34-50104.htm. Our results are robust to the inclusion of the regulation SHO

pilot firms.

http://www.sec.gov/rules/other/34-50104.htm

11

about insider stocks sales. We adopt Dow Jones’ subject categorizations. Starting with the

database described in Tetlock (2009), we have 71 news categories. However, many of these

subject codes are general codes that do not provide valuable information about the content of a

news article. For example, nearly every article in the database has the code Company News

assigned to it, in addition to a more specific news code. We remove these general codes from

our analysis to obtain a final list of subject codes that contains 39 different news categories.5

The resulting news database contains a unique firm identifier, subject codes, a dummy

variable that takes the value of one if a story was released in multiple pieces over the news day,

and two sentiment score variables that indicate whether a story contains negative words in the

headline and body of the text. The first sentiment variable is constructed using the Harvard-IV-4

dictionary as in Tetlock (2007) and Engelberg (2008) while the second sentiment measure uses

the negative word list developed by Loughran and McDonald (2009). In both cases, we

construct the sentiment score as the sum of the number of negative words in an article’s headline

and body divided by the sum of the total number of words in the headline and body.

We use the unique firm identifier to match the news data to the short sales database. The

resulting database has 1,888,868 observations over the period January 3, 2005 to July 6, 2007.

Table I contains summary statistics for the combined database. The mean number of articles per

firm-day is 1.10. However, there is substantial cross-sectional variation in this number, and

larger firms typically have more news articles on a given day. Certain news categories also

appear much more often than others. For example, the category High Yield Issuers appears

173,357 times in the database while the category 10K appears only 1,320 times. To address the

5Specifically, after computing the correlations between subject codes, we exclude subject codes if their correlation

with a more specific news category exceeds 80%. We also drop news categories that are associated with fewer than

1,000 news events over the entire sample (see Table I for the frequency of each news event in our sample).

12

potential issue of news clustering, when we conduct category specific analyses we remove

stories that are within 30 days of another story in the same category.

III. Analyses and Results

In this section we explore how short sellers differ from other traders. We begin by asking

whether short sellers respond to news before other market participants. We find that short sellers

tend to trade at the same time as other traders, and when they do not, they trade after other

traders. These results suggest that short sellers do not anticipate news. Next, we ask whether

short sellers’ trades are more profitable than other trades, consistent with a superior ability to

process news. We find consistent evidence. In a third set of tests we analyze which types of

information are associated with short sellers’ profitability. Finally, we conduct a matched

sample portfolio approach to shed additional light on the economic impact of news-based short

sales strategies.

A. Do Short Sellers Anticipate News?

One way in which short sellers may differ from other traders is in the timing of their

trades. There is some evidence that short sellers anticipate bad news announcements (e.g.,

Angel, Ferri, and Christophe (2004) and Karpoff and Lou (2009)). However, these findings

correspond to specific types of corporate events. Here we seek to shed light on short sellers’

timing behavior around all types of news events in our sample period.

13

To determine the extent of short sales timing around news events, in Figure 1 Panel A we

plot daily short sales volume (solid line), total volume (dashed line), and the ratio between the

two (dotted line) in calendar time around our universe of news events. The basic result is readily

apparent: short sellers trade when other traders do. Clearly, all traders respond to news, as there

is a significant increase in volume on the news event day and on surrounding days. However,

the ratio of short sales to total volume is nearly constant over the news period, with no significant

change in the ratio around news events. This result suggests that short sellers do not uncover and

trade on information before it becomes publicly available.

Of course, in line with the prior research above, it may be the case that short sellers

respond more to certain types of news, particularly bad news. Thus, in Panels B and C of Figure

1, we focus only on negative news events, where negative news events are defined using the

Harvard-IV-4 dictionary (see Section II.B) and the Loughran and McDonald (2009) negative

word list. The results are largely unchanged, indicating that the timing of short sellers’ response

to news does not depend on whether the news is bad.

Next, we assess whether the timing of short sellers’ trades varies by news category.

Table II presents results from a regression of short sales volume on a set of indicator variables

representing each of the news categories. Specifically, we regress daily aggregate short volume

on indicator variables that take the value one if there is a news story in a particular news category

on a given day and zero otherwise. To control for the short sellers’ response to past returns (e.g.,

Diether, Lee, and Werner (2008)), we include two lags of daily returns. The results indicate that

for the majority of news categories, short sellers respond at the same time as other traders. More

specifically, for a given news event, when we compare the coefficient estimates for regressions

on short selling after the news event to estimates for regressions on short selling before the news

14

event, we find that the coefficient estimates are largely the same. However, there are a number

of interesting exceptions. For both types of earnings news stories, Earnings and Earnings

Projections, there is more short selling after the news event than before the news event, a result

largely consistent with the findings of Angel, Ferri, and Christophe (2004). The statistically

significant estimate of 0.0049 in the t+2 specification for Earnings indicates that there is a 0.49%

increase in short selling as a percentage of total volume two days after news of this type is

reported. This late response is also apparent in news stories about joint ventures and product

distribution. In contrast, stories about leveraged buyouts show the opposite pattern. The

estimate of 0.0145 in the t-1 specification indicates that the short selling ratio increases 1.45% on

the day before news stories about leveraged buyouts, and the statistically significant coefficient

estimate on After Minus Before indicates that, relative to the two-day period before the news

event, the short selling ratio decreases 2.33% in the two-day period after the news event.

Table III presents results from a similar setup, but where the dependent variable is raw

daily short sales volume rather than short sales volume scaled by total volume. The change in

raw sales volume can be interpreted as a direct measure of the average increase or decrease in the

number of shares traded in response to news events. Our findings are qualitatively unchanged.

For instance, in the most extreme case, news days that contain Money Market News are

associated with a statistically significant increase of 291,721 additional shares sold short.

Despite the natural interpretation of these results, however, they are not as meaningful as the

scaled results discussed above. The reason is that over our sample period there is a strong

market-wide trend towards greater short volume, and thus in this analysis After Minus Before is

15

statistically significant in approximately half of the news categories.6 To summarize, in this

subsection we show that short sellers generally trade at the same time as other traders, and in

those instances in which they show different timing, short sellers tend to trade after other traders.

This suggests that short sellers’ previously documented information advantage (e.g., Boehmer,

Jones, and Zhang (2008) and Asquith, Pathak, and Ritter (2005)) does not stem from an ability to

anticipate news.

B. Do Short Sellers Have Superior Information Processing Ability?

Given our finding above, in this subsection we ask whether short sellers’ informational

advantage derives from an alternative source, namely, a superior ability to process the

information contained in publicly available news.

To answer this question as directly as possible, we begin by replicating Table IV of

Boehmer, Jones, and Zhang (2008), shown in our Table IV below. Specifically, we compute 20-

day rolling returns (i.e., t+1 to t+21) from January 3, 2005 through July 6, 2007 and we regress

these returns on the Short Volume Ratio on day t, which is defined as daily short volume divided

by total volume. The Boehmer, Jones, and Zhang (2008) result comes through strongly in these

results: in each of the specifications, Short Volume Ratio is negative and statistically significant,

indicating that when there is an increase in short sales, future prices decrease. However, given

our previous results, we might expect this pattern to be stronger among firms for which news is

released when short volume is high. To test for this effect, we include the indicator variable

6 Mean daily short volume increases from 120,910 shares in 2005 to 150,681 shares in 2007, and the increase is

statistically significant. This steady increase in short volume through time offers an explanation for the fact that

abnormal volume in Figure 1 is generally slightly above 1.

16

News Event, which takes the value one if there is news on day t and zero otherwise. We also

include contemporaneous returns as a control for the information content in the news, and we

include two days of lagged returns to control for the tendency of short sellers to trade following

recent price increases as documented by Diether, Lee, and Werner (2008). Note that because

news coverage is correlated with firm characteristics such as size and institutional ownership

(e.g., Chan (2003), Vega (2006), Engelberg (2008), and Fang and Peress (2009)), our empirical

design is meant to estimate the effect of news within firms rather than across firms. We thus

follow Skoulakis (2005) and apply the Fama-Macbeth approach to firms: we first run a time-

series regression for each firm; we then take the average of the coefficients and use the standard

deviation to estimate standard errors.

The results, shown in Table IV, provide strong evidence on the informational advantage

of short sellers. Specifically, the coefficient estimate of -0.0053 on Short-News Interaction in

Model 5 is negative and statistically significant, indicating that among stocks with high short

volume, those with news have significantly more negative future returns than those without

news.7 Even after controlling for the contemporaneous effect of returns, the coefficient on Short

Volume Ratio is still negative at the 5% level, in other words, the Boehmer, Jones, and Zhang

(2008) finding that short volume leads negative returns continues to hold. Our findings thus

provide new insight into the source of short sellers’ informational advantage. In particular, we

7 In unreported results, we add the number of negative words (a measure of the sentiment in the article) as a control

variable. This variable does not change the general magnitude or statistical significance of the results, indicating

that the findings are not driven by either very good or very bad news events.

17

find that the previously documented relation between short volume and returns is more than

twice as strong for those stocks that have a public news event.8

A drawback to using total short sales volume as a measure of short selling is that some

short sales are generated as a result of market making – to the extent that some buyers make

systematic mistakes, the corresponding short sales are simply offsetting positions, not informed

trades. Thus, with the aggregate measure of short sales volume used in Table IV, we cannot

distinguish the effect of short sales that arise in response to counterparty purchases from the

effect of shorts that arise for the purpose of gaining negative exposure. This raises the question

of whether our results in Table IV can be attributed to informed trading. To address this

concern, we take advantage of a unique feature of the data, namely, the exempt versus non-

exempt classification of trades. This classification allows us to separate shorts into market

making and non-market making (i.e., client) trades.9

Tables V and VI report the results for non-exempt and exempt trades, respectively. In

Table V the statistically significant coefficient estimate of −0.0075 in Model (5) indicates that

high short volume is a significant predictor of low future returns. Moreover, the magnitude on

short volume is 44% larger than the corresponding coefficient for total short sales volume in

Table IV, which suggests that the ability of short sales to predict future returns is particularly

8 The Boehmer, Jones, and Zhang (2008) result can be thought of as a high-frequency analog of the results in

Asquith and Muelbroek (1996) and Asquith, Pathak, and Ritter (2005). This second set of papers measures short

trading with short interest instead of short volume, and they use future returns that are measured over longer periods.

Although we would like to examine the relation between news and short sellers’ advantage in the context of these

short interest-based findings, there is an econometric challenge in making a direct comparison. Specifically, news in

our database is marked with daily time stamps, so either we would have to aggregate news to match the monthly

frequency of short interest or we would have to throw out much of our news data. It is not clear how a reduction in

the frequency of the news variable would change expectations about the short positions. 9 Anecdotal evidence suggests that the exemption is sometimes abused, but only in one direction: trades may be

inappropriately marked as exempt when they are not. Since the exemption removes potential restrictions, it is

unlikely that exempt trades would ever be inappropriately marked as non-exempt. In other words, exempt trades

may include client trades, but non-exempt trades are unlikely to include market maker trades.

18

strong for non-market-making trades. We also see that the short-news interaction estimate of

−0.0056 is significantly negative, indicating that non-market makers’ shorts are 75% (i.e.,

−0.0056/−0.0075) more profitable in the presence of news events than at other times. In

contrast, the results in Table VI indicate that market makers’ trades are not particularly well

informed: the short volume ratio loads as a positive predictor of price, indicating that market

makers’ shorts are actually associated with positive future returns on average.10

Further, there is

no differential effect of market makers’ trading in the presence of news.

Overall, the evidence in this subsection suggests that the most informed short sales are

made for the purpose of gaining negative exposure, and that these trades are particularly well

informed in the presence of recent news events.

C. Price Responses by News Category

In an extension to the above analysis, in this subsection we ask whether short sellers’

information processing ability is uniformly strong across news categories. To get at this

question, we repeat the analysis in Table IV separately for each news category. Specifically, for

each news category we run a regression in which the dependent variable is the compound return

from the first to the twentieth trading day after the news event, and the main independent

variable is short vol / market vol, which is the amount of short selling relative to total volume on

the day of the news event. Since the type of news (good or bad) may have some effect on future

10

Even though the magnitudes of the coefficients vary between Table V and Table VI, the economic impact is of the

same order of magnitude. Specifically, a one standard deviation increase in short volume among non-exempt trades

leads to a 0.200% decrease in future returns, while a one standard deviation increase in short volume among non-

exempt trades is associated with a 0.154% increase in future returns.

19

returns (e.g., Bernard & Thomas (1989)), we attempt to control for whether the news is good or

bad by including the event-day return on the right-hand side of the specification.

The results, shown in Table VII, indicate that short sellers have some ability to identify

trades that are likely to be profitable around certain news events. Specifically, we find that the

coefficient estimates on short vol / market vol are significantly negative for 12 of the 39 news

categories; of these, five are statistically significant at the 1% level (Corporate Restructurings,

Earnings, Earnings Projections, New Products & Services, and Stock Ownership), and nine are

significant at the 5% level (Corporate Restructurings, Divestitures or Asset Sales, Earnings,

Earnings Projections, Initial Public Offerings, Management Issues, New Products & Services,

Research and Development, and Stock Ownership). As a further test of statistical significance,

we conduct a Fisher test of combined probability to determine whether the cross-sectional

distribution of the p-values from each regression differs significantly from a uniform zero-one

distribution. The Fisher test rejects this null at the 1% level of significance across all news

categories, suggesting that the coefficient on short volume is statistically different from zero for

the cross-section. We also find that many of the categories that are statistically significant are

the same categories identified in Subsection A as the categories in which short sellers’ trades are

measurably later than other investors’ trades (e.g., Earnings, and Earnings Projections).

Taken together, these results indicate that when short selling predicts future returns, short

sellers appear to be making profitable trades. This evidence lends further support to the idea that

short sellers’ informational advantage stems from superior ability to process publicly available

information.

20

D. Matched Sample Portfolio Approach

So far we provide evidence that short selling is more informative on news event days. In

this subsection we shed light on the economic impact of news-based short selling strategies using

a portfolio approach. This approach recognizes that the presence of news is likely correlated

with firm characteristics and that certain categories of news may be more relevant for some firms

than for others. In other words, since news is strongly related to several firm characteristics, we

cannot simply sort on news. Moreover, news coverage is highly persistent: firms that have many

news articles in the Dow Jones archive in one year are likely to have many articles in following

years. Thus, in order to conclude that news, rather than a particular firm characteristic, is driving

the differential returns we observe we need to compare two firms that are identical apart from the

fact that one firm has a news event while the other firm does not. We do this using a matched

sample portfolio approach.

Our approach is based on forming portfolios of stocks around news events. Because

previous research indicates that firm characteristics may affect future returns, we implement a

control sample methodology to control for these previously documented effects. Specifically, for

every stock with a news event, we identify a control stock that is the closest match in the

following four dimensions: bid-ask-spread, institutional ownership, market capitalization, and

number of news events over the previous month. We match by selecting the stock that

minimizes the sum of the rank differences in each of these categories. Furthermore, to eliminate

potentially contaminating competitive effects (e.g., Slovin, Sushka, and Bendeck (1991), Chen,

21

Ho, and Ik (2005), and Hsu, Reed, and Rocholl (2009)), we require that control firms and sample

firms be members of different Fama-French 48 industries.11

The analysis yields results for each of the 39 news categories. Figure 2 presents the

results for three categories as examples. In the Dividends category, we see that among firms

with dividend news, firms with high short volume have significantly lower returns than firms

with low short volume. This difference is approximately 4.39% at the one-year point. In

contrast, the control sample shows similar returns across high short volume stocks and low short

volume stocks.

Table VIII summarizes the detailed results of this analysis. The economic significance of

news-based short sales becomes apparent when we compare differences in portfolio returns. For

example, if an investor were to sell a portfolio of stocks with high short selling and buy a

portfolio of stocks with low short selling on the day that Product Distribution news is released,

that investor would earn an excess annual return of 6.50%. The same strategy for a matched

portfolio of no-news stocks would return 5.74% over the period, yielding a difference of

12.24% annually between the two strategies. In fact, this strategy yields positive excess returns

in 34 out of our 39 news categories, with some news categories yielding annualized excess

returns of over 10%. The statistically significant 2.89% return for the mean excess return

difference indicates that not only do short sellers have a significant advantage over other traders,

but their advantage comes largely from their ability to process and trade on news events.

To summarize, this analysis shows that the inverse relation between short volume and

future returns is strongest around news events, whereas during non-news events this relation may

11

In unreported results, we use the Fama-French 12 industry classifications instead of the Fama-French 48

classifications. The results are not qualitatively different.

22

be insignificant or even go in the other direction. These results lend additional support to our

main finding that the previously documented informational advantage of short sellers is driven in

large part by short sellers’ superior ability to process information contained in publicly available

news.

IV. Conclusion

Previous research documents that short sellers are informed traders (e.g., Boehmer, Jones,

and Zhang (2008) and Asquith, Pathak, and Ritter (2005)). Yet we know little about the source

of short sellers’ informational advantage. This paper seeks to fill this gap by investigating the

following questions: To what extent are short sellers able to anticipate news events? Are short

sellers better able to process and react to news? And, are short sellers’ trades particularly

profitable around specific categories of news? To address these questions, we combine a

database of all public news events in the U.S. with a database of short sale trades over the same

sample period.

We find that, in general, short sellers trade at the same time as other traders. Specifically,

the ratio of short sales to total volume is nearly constant over news periods, with no significant

change in the ratio around news events. However, we do find some differences between the

timing of short sellers’ trades and the overall market: for news stories about analysts’ comments

and ratings, earnings, earnings projections, joint ventures, and product distribution, there is a

significant increase in short selling after the news story. This finding suggests that like other

traders, short sellers trade on publicly available information, and hence their informational

advantage is not due to an ability to uncover or anticipate information before it becomes public.

23

Given the result that short sellers’ advantage is not due to timing, we next ask if it could

be due to superior ability to process the information available in public news stories. We find

supportive evidence. In particular, we find that across all types of news, short selling predicts

future returns: even after controlling for the unconditional relation between short selling and

news (e.g., Boehmer, Jones, and Zhang (2008)), short selling’s predicative effect on future

returns is more than twice as strong in the presence of news. This result is not a reflection of

persistent mistakes by buyers, that is, the most informed short sales are not from market makers

but from clients, and these client shorts are particularly well informed in the presence of news.

We also find that this predicative effect is strongest for nine categories of news (Corporate

Restructurings, Divestitures or Asset Sales, Earnings, Earnings Projections, Initial Public

Offerings, Management Issues, New Products & Services, Research and Development, and Stock

Ownership), and that many of these categories are the same categories for which short sellers’

timing follows the overall market. Finally, recognizing that the presence of news is likely

correlated with firm size and that certain categories of news may be more relevant for some firms

than for others, we conduct an experiment in which each firm’s response to a news event is

matched by a control firm’s response on the same day. We find that across all news categories,

short sellers’ advantage in predicting returns is concentrated in firms with news.

In sum, we show that, on average, short sellers’ advantage is not due to an ability to

influence the public’s perception of value, as recently suggested by the Securities and Exchange

Commission.12

Rather, we find that short sellers generally trade when other traders do, and to

the extent that the timing of their trades differs, short sellers actually trade after other traders.

We further find that short sellers’ ability to predict future negative returns is concentrated around

12

Short sellers were accused of “distort and short” schemes in “What the SEC Really Did on Short Selling” by

Chairman Christopher Cox, 24 July 2008, The Wall Street Journal.

24

news events. Thus, by connecting short sellers’ trading patterns with news releases, we show

that short sellers’ trading advantage derives primarily from their superior ability to analyze

publicly available information.

The findings of this paper shed light on the broader debate about the informational effects

of news announcements. While several papers argue that public news announcements reduce

information asymmetry (e.g., Korajczyk, Lucas, and McDonald (1991), Kacperczyk and Seru

(2007)), others have recognized that public news events lead to differential interpretations by

traders (Kandel and Pearson (1995)) based on the skill of those traders. Rubenstein (1993) puts

it succinctly: “In real life, differences in consumer behavior are often attributed to varying

intelligence and ability to process information. Agents reading the same morning newspapers

with the same stock price lists will interpret the information differently.” This view explains not

only why volume is high around news events (Kandel and Pearson (1995)) but also why some

papers find return predictability from “soft” information in news announcements (e.g., Engelberg

(2008), Demers and Vega (2008)). Specifically, public information events present profitable

trading opportunities for skilled information processors. Tetlock (2009) weighs in on this

question by modeling liquidity shocks and news and finds empirical evidence suggesting that

public information plays a key role in informing a subset of investors. By focusing specifically

on short sellers, this paper has a unique ability to shed light on the trade-level evidence in this

debate for two reasons: first, a number of papers show that short sellers are informed (e.g.,

Asquith and Meulbroek (1995), Diamond and Verrecchia (1987)); second, short sellers’ trades

are among the few classes of trades that are uniquely identified. Our finding that short sellers’

trades are more than twice as profitable in the presence of recent news is strong evidence in favor

of the idea that news presents profitable trading opportunities for skilled information processors.

25

REFERENCES

Antweiler, W. and M. Frank, 2006, Do U.S. stock markets typically overreact to corporate news

stories?, Working Paper, University of British Columbia.

Asquith, P., and L. Meulbroek, 1995, An empirical investigation of short interest, Unpublished

Working Paper, M.I.T.

Asquith, P., P. Pathak, and J. Ritter, 2005, Short Interest, Institutional Ownership, and Stock

Returns, Journal of Financial Economics 78, 243-276.

Aitken, M., A. Frino, M. McCorry, and P. Swan, 1998, Short sales are almost instantaneously

bad news: Evidence from the Australian Stock Exchange, Journal of Finance 53, 2205-

2223.

Bernard, V. and J. Thomas, 1989, Post-earnings-announcement drift: delayed price response or

risk premium?, Journal of Accounting Research 27, 1-36.

Boehmer, E., C. Jones and X. Zhang, 2008, Which Shorts are Informed?, Journal of Finance 63,

491-527.

Chakrabarty, Bidisha, and Andriy Shkilko, 2008, Information Leakages in Financial Markets:

Evidence from Shorting around Insider Sales, Working Paper.

Chan, W., 2003, Stock price reaction to news and no-news: drift and reversal after headlines,

Journal of Financial Economics 70, 223-260.

Chen, S., K. W. Ho, and K. H. Ik, 2005, The Wealth Effect of New Product Introductions on

Industry Rivals, Journal of Business 78, 969-996.

Christophe, S., M. Ferri, and J. Angel, 2004, Short-Selling Prior to Earnings Announcements,

Journal of Finance 59, 1845-1875.

Daske, H. S. Richardson, and A. Tuma, 2005, Do Short Sale Transactions Precede Bad News

Events?, Working Paper.

Diamond, D., and R. Verrecchia, 1987, “Constraints on short-selling and asset price adjustment

to private information,” Journal of Financial Economics, 18, 277–311.

Diether, K., K. Lee, and I. Werner, 2008, Short-sale Strategies and Return Predictability, Review

of Financial Studies 22, 575-607.

Edwards, A. and K. Hanley, 2008, Short Selling in Initial Public Offerings, Working Paper.

Engelberg, J., 2008, Costly Information Processing: Evidence from Earnings Announcements,

Working Paper, University of North Carolina.

26

Evans, Richard, Chris Geczy, David Musto and Adam Reed, 2009, “Failure is an Option:

Impediments to Short-Selling and Options Prices”, The Review of Financial Studies

22(5), 2009.

Fang, L. and J. Peress, 2009, Media Coverage and the Cross-Section of Stock Returns,

Forthcoming in the Journal of Finance.

Feldman, Ronen, Suresh Govindaraj, Joshua Livnat, and Benjamin Segal, 2008, The incremental

information content of tone change in management discussion and analysis, Working

paper, INSEAD.

Fox, Merritt B., Lawrence Glosten, and Paul Tetlock, 2009, Short Selling and the News: A

Preliminary Report on an Empirical Study, Working Paper.

Gervais, S., R. Kaniel, and D. Mingelgrin, 2001, The High-Volume Return Premium, Journal of

Finance 56, 877-919.

Hsu, H., A. Reed, and J. Rocholl, 2009, The new game in town: competitive effects of IPOs,

Forthcoming in the Journal of Finance.

Kacperczyk, M. and A. Seru, 2007, Fund Manager Use of Public Information: New Evidence on

Managerial Skills, Journal of Finance 62, 485-528.

Kandel, Eugene, and Neil D. Pearson, 1995, Differential interpretation of public signals andtrade

in speculative markets, Journal of Political Economy 103, 831-872.

Karpoff, J. and X. Lou, 2009, Short sellers and financial misconduct, Working Paper.

Loughran, Tim and Bill McDonald, 2009, When is a Liability not a Liability? Textual Analysis,

Dictionaries, and 10-Ks, Working Paper, University of Notre Dame.

Nagel, S., 2005, Short sales, institutional investors and the cross-section of stock returns, Journal

of Financial Economics 78, 277-309.

Rubinstein, Ariel, 1993, On Price Recognition and Computational Complexity in a Monopolistic

Model, Journal of Political Economy 101, 473-84.

Senchack, A.J., Jr., and L. Starks, 1993, Short-sale restrictions and market reaction to short-

interest announcements, Journal of Financial and Quantitative Analysis 28, 177-194.

Skoulakis, G., 2005, Assessment of Asset-Pricing Models using Cross-Sectional Regressions,

Working paper, Northwestern University.

Slovin, M. B., M. E. Sushka, and Y. M. Bendeck, 1991, The Intra-Industry Effects of Going-

Private Transactions, Journal of Finance 46, 1537-1550.

27

Tetlock, Paul C., 2007, Giving content to investor sentiment: The role of media in the stock

market, Journal of Finance 62, 1139-1168.

Tetlock, Paul C., 2008, All the News That’s Fit to Reprint: Do Investors React to Stale

Information?, Working Paper.

Tetlock, Paul C., 2009, Does Public Financial News Resolve Asymmetric Information?,

Working Paper.

Tetlock, Paul C., M. Saar-Tsechansky, and S. Macskassy, 2008, More Than Words: Quantifying

Language to Measure Firms' Fundamentals, Journal of Finance 63, 1437-1467.

Vega, C., 2006, Stock price reaction to public and private information, Journal of Financial

Economics 82, 103-133.

28

Figure 1

Volume around News Events

Figure 1 displays short volume, total volume, and the ratio of short volume to total volume for

the 15 days before and after news events. Short volume and total volume are scaled by their

mean values over the period -16 to -30. Panel A displays volume around all news events.

Panels B and C display volume for negative news events only. Version 1 (Panel B) uses a

negative sentiment variable that was constructed using the Harvard-IV-4 Dictionary as in

Tetlock (2007) and Engelberg (2008) while version 2 (Panel C) uses a sentiment measure that

employs the negative word list developed by Loughran and McDonald (2009).

Panel A: All News Events

Panel B: Negative News Events – version 1

Panel C: Negative News Events – version 2

0.0000

0.2000

0.4000

0.6000

0.8000

1.0000

1.2000

1.4000

1.6000

1.8000

-15 -14 -13 -12 -11 -10 -9 -8 -7 -6 -5 -4 -3 -2 -1 0 +1 +2 +3 +4 +5 +6 +7 +8 +9 +10+11+12+13+14+15

Vo

lum

e

All News Events

Short Volume Total Volume Short Vol / Total Vol

0.0000

0.2000

0.4000

0.6000

0.8000

1.0000

1.2000

1.4000

1.6000

-15 -14 -13 -12 -11 -10 -9 -8 -7 -6 -5 -4 -3 -2 -1 0 +1 +2 +3 +4 +5 +6 +7 +8 +9 +10+11+12+13+14+15

Vo

lum

e

Negative News Events - Version 1


0.0000

0.2000

0.4000

0.6000

0.8000

1.0000

1.2000

1.4000

1.6000

-15 -14 -13 -12 -11 -10 -9 -8 -7 -6 -5 -4 -3 -2 -1 0 +1 +2 +3 +4 +5 +6 +7 +8 +9 +10+11+12+13+14+15

Vo

lum

e

Negative News Events - Version 2


29

Figure 2

Example Short Volume Portfolio Returns following News Events

Figure 2 displays buy and hold portfolio returns for a 12 month period following news events.

Each day for each news event, two portfolios are formed: the first portfolio consists of those

firms that had a specific news event and had low short volume as a percentage of total volume;

the second portfolio consists of those that had the news event and had high short volume as a

percentage of total volume. We then form control portfolios using a sample of firms that did not

experience a news event but were similar in terms of bid-ask-spread, institutional ownership,

market capitalization, and the number of news events over the previous month. The detailed

results are shown in Table VIII and three example results are shown below. Panel A displays

portfolio returns following dividend news and the returns for the matched control sample. Panel

B displays portfolio returns following earnings news and the associated control returns and Panel

C contains returns following news about insider stock sales and the associated control returns.

Panel A: Dividends

News Sample Control Sample

Panel B: Earnings


Panel C: Insider Stock Sales


-1.0%

0.0%

1.0%

2.0%

3.0%

4.0%

5.0%

0 1 2 3 4 5 6 7 8 9 10 11 12

Bu

y a

nd

Ho

ld R

etu

rn

Months after Portfolio Formation

Low Short Volume High Short Volume

-1.0%

0.0%

1.0%

2.0%

3.0%

4.0%

5.0%

0 1 2 3 4 5 6 7 8 9 10 11 12

Bu

y a

nd

Ho

ld R

etu

rn



-1.0%

0.0%

1.0%

2.0%

3.0%

4.0%

5.0%

6.0%

7.0%

0 1 2 3 4 5 6 7 8 9 10 11 12

Bu

y a

nd

Ho

ld R

etu

rn



-1.0%

0.0%

1.0%

2.0%

3.0%

4.0%

5.0%

6.0%

7.0%

8.0%

0 1 2 3 4 5 6 7 8 9 10 11 12

Bu

y a

nd

Ho

ld R

etu

rn



-3.0%

-2.0%

-1.0%

0.0%

1.0%

2.0%

3.0%

4.0%

0 1 2 3 4 5 6 7 8 9 10 11 12

Bu

y a

nd

Ho

ld R

etu

rn



0.0%

1.0%

2.0%

3.0%

4.0%

5.0%

6.0%

7.0%

8.0%

0 1 2 3 4 5 6 7 8 9 10 11 12

Bu

y a

nd

Ho

ld R

etu

rn



30

Table I

Summary Statistics

The database has 1,888,868 observations over the period January 3, 2005 through July 6, 2007.

Panel A provides summary statistics at the firm level. News articles may be reissued throughout

the day as more information becomes available; in such situations we consider all of the related

article updates to be one unique news event and we keep track of the number of articles that are

rolled-into this unique news event. News Articles per Firm-Day is a count of all news articles

including reissued (updated) articles while Unique News Events per Firm-Day is a count of the

unique stories, excluding subsequent updates to an article. Short Vol. / Total Vol. is the short

volume from the NYSE TAQ Regulation SHO database as a percentage of total volume; Exempt

and Non-Exempt denote market maker short sales (exempt) from non-market maker short sales

(non-exempt), see section II.A for details. Market Capitalization is from CRSP. Panel B

contains summary statistics on the frequency of each news category in the database as well as

the mean number of negative words as a percentage of total words in the headline and body text

of each article. News articles may be classified into more than one category and we adopt two

methods for counting the number of negative words in the headline and body text of each article:

Version 1 uses the Harvard-IV-4 Dictionary as in Tetlock (2007) and Engelberg (2008) while

Version 2 uses the negative word list developed by Loughran and McDonald (2009).

Panel A – Firm Level Statistics Mean Median 1

st

Percentile

99

th

Percentile

Standard

Deviation

News Articles per Firm-Day 1.10 0.00 0.00 18.00 4.09

Unique News Events per Firm-Day 0.81 0.00 0.00 12.00 2.77

Short Vol. / Total Vol. 19.60% 17.47% 0.52% 62.46% 27.22%

Short Vol. / Total Vol. – Exempt 3.61% 1.44% 0.01% 32.26% 8.11%

Short Vol. / Total Vol. – Non-exempt 17.63% 15.66% 0.39% 55.37% 26.76%

Market Capitalization ($ mm) $5,856 $1,228 $32 $80,360 $19,329

Panel B – News Categories N

Mean Negative Headline

Words (% of total)

Mean Negative Body

Text Words (% of total)

Version 1 Version 2 Version 1 Version 2

10K 1,320 6.65% 4.07% 3.35% 1.79%

8K 10,803 4.83% 1.80% 2.91% 1.10%

Acquisitions, Mergers, Takeovers 56,993 5.09% 1.36% 3.18% 1.00%

Analysts' Comments & Ratings 49,508 5.00% 2.12% 3.16% 1.08%

Annual Meetings 4,041 6.63% 1.32% 3.01% 0.93%

Antitrust News 5,217 7.40% 3.88% 3.99% 1.72%

Bankruptcy-Related Filings 6,258 6.88% 3.30% 4.47% 1.99%

Bond Ratings & Comments 15,343 6.39% 1.62% 3.29% 1.15%

Buybacks 6,269 4.94% 1.03% 3.01% 0.93%

Contracts, Defense 4,734 6.03% 1.85% 3.24% 1.07%

Contracts, Government (not defense) 3,321 5.92% 1.44% 3.20% 1.01%

31

Table I (continued)

Panel B – News Categories N

Mean Negative Headline

Words (% of total)

Mean Negative Body

Text Words (% of total)

Version 1 Version 2 Version 1 Version 2

Contracts, Nongovernment 19,102 5.96% 1.16% 2.93% 0.82%

Corporate Governance 4,981 7.00% 2.49% 3.96% 1.55%

Corporate Restructurings 5,631 6.04% 2.16% 3.78% 1.43%

Divestitures or Asset Sales 11,587 4.85% 1.25% 3.21% 1.03%

Dividend News 24,731 7.64% 0.60% 2.64% 0.54%

Earnings 40,705 5.84% 1.09% 3.12% 0.89%

Earnings Projections 37,432 5.57% 1.39% 3.26% 0.99%

Financing Agreements 6,919 5.01% 1.27% 2.95% 0.98%

High-Yield Issuers 173,357 5.24% 1.68% 2.92% 0.90%

Initial Public Offerings 9,351 3.70% 1.01% 2.59% 0.76%

Insider Stock Buys 21,489 1.65% 0.45% 1.15% 0.40%

Insider Stock Sells 54,868 1.68% 1.29% 1.34% 0.34%

Joint Ventures 9,081 5.71% 1.20% 3.03% 0.88%

Labor Issues 12,376 7.12% 2.49% 3.89% 1.44%

Lawsuits 15,351 7.23% 3.85% 4.37% 2.40%

Leveraged Buyouts 2,286 6.07% 1.46% 3.79% 1.16%

Management Issues 13,840 5.58% 1.71% 3.38% 1.18%

Market News 14,068 6.16% 2.01% 3.77% 1.31%

Money Market News 1,298 6.61% 1.99% 4.00% 1.37%

New Products & Services 24,583 7.08% 1.24% 3.02% 0.77%

Personnel Appointments 29,994 6.71% 1.36% 3.19% 0.86%

Point of View 17,316 6.09% 1.84% 3.88% 1.26%

Product Distribution 2,440 5.90% 1.52% 3.21% 0.99%

Research & Development 5,323 6.80% 1.88% 3.76% 1.21%

Spinoffs 1,874 3.89% 1.17% 2.68% 0.85%

Stock Options 5,679 5.42% 1.95% 3.38% 1.27%

Stock Ownership 25,567 2.69% 0.70% 2.38% 0.34%

Stock Splits 2,230 4.23% 0.56% 2.13% 0.35%

32

Table II

Regression Analysis of Short Volume Ratio around News Events

Table II contains the results of six regressions of short sales volume on a set of indicator variables representing stories in each of the

news categories. Specifically, the dependent variable is aggregate short volume as a percentage of total volume and the independent

variables are indicator variables that take the value one if there is a news story in a particular news category and zero otherwise. We

vary the timing of the dependent variable relative to the news event in order to examine short volume changes around news. For

example, t-2 indicates that the dependent variable is observed two days prior to the news event. For After Minus Before the dependent

variable is the difference in the short volume ratio between dates t+2 and t-2. To control for the documented response of short sellers

to past returns, we include two lags of daily returns. *** indicates significance at the 1% level, ** indicates significance at the 5%

level, and * indicates significance at the 10% level.

After

Event Time of the Dependent Variable Minus

News Events t-2 t-1 t=0 t+1 t+2 Before

Mean of the fixed effects 0.1823*** 0.1855*** 0.1844*** 0.1835*** 0.185*** -0.0102

Return (1 day lag) 0.3878*** 0.3854*** 0.3882*** 0.3893*** 0.3904*** 0.5749***

Return (2 day lag) 0.2715*** 0.2729*** 0.2735*** 0.2738*** 0.2744*** -0.4506***

10K -0.0007 -0.0036 -0.0039 -0.0031 -0.0018 0.0046

8K -0.0011 -0.0019 -0.0003 -0.0003 -0.0003 0.0012

Acquisitions, Mergers, Takeovers 0.0023 0.0006 0.0001 0.0000 0.0006 0.0268

Analysts' Comments & Ratings of Stocks 0.0017 0.0023 0.0123*** 0.006*** 0.0063*** 0.0078**

Annual Meetings -0.0084** -0.0022 -0.0016 -0.0037 -0.0013 -0.0016

Antitrust News -0.0041 -0.0059 -0.0028 -0.0046 -0.0053 0.0005

Bankruptcy-Related Filings 0.0035 0.0011 -0.0004 -0.0035 -0.0013 -0.0124

Bond Ratings & Comments 0.0005 -0.0005 -0.0005 -0.0018 -0.0012 0.0801

Buybacks -0.0038 -0.0052* -0.0017 -0.0041 -0.003 0.0018

Contracts, Defense -0.0021 -0.0028 0.0078 0.0025 0.0009 0.0092

Contracts, Government (not defense) -0.0059 -0.0084* -0.0039 -0.0010 -0.0015 0.0098

Contracts, Nongovernment 0.0008 0.0014 -0.0012 0.0003 -0.0006 -0.0024

Corporate Governance -0.0008 -0.0026 -0.0004 -0.0001 0.0008 0.0037

Corporate Restructurings 0.0033 0.0021 0.0004 -0.0024 -0.0001 0.0044

Divestitures or Asset Sales -0.0006 -0.0011 -0.0010 -0.0018 0.0000 0.0005

Dividend News 0.0015 -0.0034** -0.0005 0.0003 0.0005 0.0048*

Earnings 0.0008 -0.0027* -0.0025 0.0028* 0.0049*** 0.0132***

33

Table II (continued) After



Earnings Projections -0.0036* 0.0004 -0.0007 0.0031 0.0025 0.007**

Financing Agreements -0.0017 -0.0035 -0.0038 -0.0006 -0.0030 0.0013

High-Yield Issuers -0.0004 0.0019 0.0057** 0.0035 0.0024 0.0046

Initial Public Offerings -0.0030 0.0015 0.0009 -0.0025 -0.0009 -0.0038

Insider Stock Buys -0.0024 -0.0029 -0.0021 -0.0014 -0.0023 -0.0002

Insider Stock Sells -0.0058** -0.0059** -0.0066** -0.0040 -0.0011 0.0069

Joint Ventures -0.0046 -0.0051* -0.0017 0.0028 -0.0004 0.0123**

Labor Issues -0.0013 -0.0006 -0.0016 0.0009 -0.0008 0.0033

Lawsuits 0.0053* -0.0014 -0.0053* -0.0003 0.0011 0.0085

Leveraged Buyouts -0.0031 0.0145*** -0.0119** -0.0025 -0.0063 -0.0233**

Management Issues -0.0036 -0.0024 -0.0047* 0.0000 -0.0020 -0.0048

Market News -0.0030 -0.0037 -0.0057* -0.0042 -0.0056* -0.0034

Money Market News -0.0018 -0.0019 -0.0105 0.0006 -0.0083 0.0025

New Products & Services 0.0021 -0.0025 -0.0020 0.0001 0.0017 0.0013

Personnel Appointments -0.0012 0.0024 -0.0002 0.0013 0.0034* 0.0043

Point of View -0.0011 0.0016 -0.0038 -0.0028 -0.002 0.0025

Product Distribution -0.0054 -0.0035 -0.0042 -0.0012 -0.0028 0.0055**

Research & Development -0.0019 -0.0003 -0.0058 -0.0051 -0.0021 -0.0010

Spinoffs -0.0057 -0.0037 -0.0069 -0.0059 -0.0087 -0.0044

Stock Options -0.0030 -0.0031 -0.0040 -0.0038 -0.0035 -0.0076

Stock Ownership -0.0045*** -0.0042*** -0.0041*** -0.0023* -0.0010 0.0041

Stock Splits 0.0017 0.0000 0.0047 0.0018 0.0102*** -0.0032

34

Table III

Regression Analysis of Raw Short Volume around News Events

Table III contains the results of six regressions of short sales volume on a set of indicator variables representing stories in each of the

news categories. Specifically, the dependent variable is the raw aggregate short volume (not scaled) and the independent variables are

indicator variables that take the value one if there is a news story in a particular news category and zero otherwise. We vary the

timing of the dependent variable relative to the news event in order to examine short volume changes around news. For example, t-2

indicates that the dependent variable is observed two days prior to the news event. For After Minus Before the dependent variable is

the difference in the short volume ratio between dates t+2 and t-2. To control for the documented response of short sellers to past

returns, we include two lags of daily returns. *** indicates significance at the 1% level, ** indicates significance at the 5% level, and

* indicates significance at the 10% level.

After



Mean of the fixed effects 141,452 137,444 128,403 120,962* 144,973 -123,750***

Return (1 day lag) 351,936*** 353,335*** 335,762*** 331,588*** 342,662*** 549,513***

Return (2 day lag) 160,030*** 160,302*** 157,685*** 152,383*** 151,151*** -761,974***

10K -23,762*** -28,923*** 23,824*** 24,974*** 1,230 11,559

8K 7,775** 7,171* 14,523*** 14,654*** 3,666 72,636***

Acquisitions, Mergers, Takeovers 1,946 -1,256 11,297*** 8,393** 6,162* -119,140**

Analysts' Comments & Ratings of Stocks 4,658 31,158*** 77,550*** 26,301*** 11,636*** 5,678

Annual Meetings 3,465 15,713*** 20,810*** 12,863** 8,636 -20,271

Antitrust News 3,173 -7,773 26,145*** 3,301 10,656* 65,438***

Bankruptcy-Related Filings 15,192** 22,822*** 90,670*** 42,514*** 42,438*** 47,730***

Bond Ratings & Comments 7,069* 29,597*** 77,503*** 28,790*** 5,213 -141,097

Buybacks 19,459*** 21,194*** 92,991*** 53,651*** 19,574*** 33,025***

Contracts, Defense -31,532*** -27,023** -11,273 -10,146 -10,117 54,981**

Contracts, Government (not defense) -13,446 -18,100** -3,401 1,168 -3,031 27,580

Contracts, Nongovernment 4,944 2,579 14,985*** 12,979*** 4,351 8,176

Corporate Governance 20,671*** 29,455*** 44,600*** 27,269*** 13,420** 2,389

Corporate Restructurings 806 10,367* 57,111*** 37,131*** -7,468 7,584

Divestitures or Asset Sales 4,527 9,350** 38,324*** 10,396** 2,118 -3,634

Dividend News 6,770*** 2,951 12,491*** 17,614*** 8,102*** 16,521***

Earnings -8,809*** 1,668 63,417*** 65,849*** 25,914*** 101,013***

35

Table III (continued)

After



Earnings Projections -4,336 3,161 53,246*** 44,373*** 16,240*** 63,330***

Financing Agreements 6,598 2,358 29,615*** 13,609*** -147 2,576

High-Yield Issuers -3,626 -2,538 37,108*** 4,242 1,661 11,645

Initial Public Offerings 17,781*** 53,697*** 60,989*** 19,461*** 9,143 -36,536**

Insider Stock Buys 14,979*** 17,307*** 8,147** 4,510 1,752 -26,275***

Insider Stock Sells 18,020*** 5,991 -1,482 4,197 6,098 -10,966

Joint Ventures 10,298** -528 8,540* 10,657** 4,212 9,445

Labor Issues 6,204 14,395*** 65,368*** 32,374*** 22,333*** 37,023***

Lawsuits 4,035 14,461*** 16,622*** 15,239*** 1,014 79,158

Leveraged Buyouts 25,052*** 56,619*** 172,693*** 107,072*** 46,789*** 55,578***

Management Issues -4,362 1,217 -6,665 -820 -4,285 85,346***

Market News 6,757 22,521*** 199,417*** 87,374*** 22,841*** -551

Money Market News 89,479*** 135,472*** 291,721*** 133,469*** 64,114*** -172

New Products & Services -2,304 -6,230 -3,791 2,751 3,359 65,542***

Personnel Appointments 2,044 -711 13,598*** 6,932** 3,177 -6,578

Point of View 19,505*** 24,110*** 30,348*** 15,518** 12,959* 13,839

Product Distribution -14,420** 6,252 6,156 6,571 -17,414** 4,708

Research & Development 21,134** -6,816 3,107 -9,898 -3,898 -8,813

Spinoffs -1,126 77,486*** 66,522*** 42,935*** 29,173*** -7,178

Stock Options 6,848 17,998*** 97,709*** 47,563*** 42,051*** 16,458

Stock Ownership 2,128 7,942*** 9,474*** 8,505*** 3,436 7,098

Stock Splits -4,078 27,491*** 14,689** 433 -3,889 -29,274

36

Table IV

Cross-Sectional Relation between Returns, Short Sales, and News

Table IV contains the results from Fama-MacBeth (1973) type regressions using daily

observations over the period January 3, 2005 through July 6, 2007. The regressions are done

firm by firm, and the dependent variable is the buy and hold (compound) return over the next 20

trading days. Panel A is calculated using raw returns as the dependent variable while Panel B

uses characteristic adjusted returns as in Daniel, Grinblatt, Titman, and Wermers (1997),

however we omit the book to market factor due to missing Compustat data for some firms. The

Short Volume Ratio is daily short volume / total volume. News Event is an indicator variable

that takes the value one if a news event occurs for a particular stock, and Short-News Interaction

is the product of Short Volume Ratio and the News Event indicator. Returnt=0 is the return on

each stock on the day that short volume and news are observed. T-statistics are below the

parameter estimates in italics and are calculated using Newest-West (1987) standard errors with

20 lags. *** indicates significance at the 1% level, ** indicates significance at the 5% level, and

* indicates significance at the 10% level.

Panel A: Raw Returns

Model

(1) (2) (3) (4) (5)

Intercept 0.0171*** 0.0173*** 0.0171*** 0.0169*** 0.0168***

(4.70) (4.78) (4.73) (4.82) (4.99)

Short Volume Ratio -0.0053** -0.0053** -0.0044* -0.0047** -0.0052**

(-2.19) (-2.20) (-1.86) (-2.01) (-2.24)

News Event -0.0010 0.0000 0.0000 0.0000

(-1.37) (0.02) (0.05) (0.06)

Short – News Interaction -0.0050** -0.0053*** -0.0053***

(-2.41) (-2.62) (-2.72)

Returnt=0 0.0213 0.0227

(1.29) (1.35)

Returnt=-1 0.0287*

(1.86)

Returnt=-2 0.0356**

(2.38)

Panel B: DGTW Returns

Intercept 0.0054*** 0.0056*** 0.0054*** 0.0052*** 0.0051***

(5.32) (5.16) (4.97) (4.62) (4.07)

Short Volume Ratio -0.0069*** -0.0068*** -0.0058*** -0.0061*** -0.0065***

(-3.27) (-3.27) (-2.85) (-3.00) (-3.23)

News Event -0.0009* 0.0003 0.0003 0.0003

(-1.72) (0.35) (0.42) (0.44)

Short – News Interaction -0.0059*** -0.0063*** -0.0063***

(-2.75) (-2.98) (-3.12)

Returnt=0 0.0216 0.0226

(1.37) (1.41)

Returnt=-1 0.0303**

(2.03)

Returnt=-2 0.0333**

(2.33)

37

Table V

Cross-Sectional Relation between Returns, Short Sales, and News for Non-Exempt Trades

Table V contains the results from Fama-MacBeth (1973) type regressions using daily

observations over the period January 3, 2005 through July 6, 2007. The sample only includes

those short sales transactions that were classified as non-exempt as discussed in Section II.A of

the text. The regressions are done firm by firm, and the dependent variable is the buy and hold

(compound) return over the next 20 trading days. Panel A is calculated using raw returns as the

dependent variable while Panel B uses characteristic adjusted returns as in Daniel, Grinblatt,

Titman, and Wermers (1997), however we omit the book to market factor due to missing

Compustat data for some firms. The Short Volume Ratio is daily short volume / total volume.

News Event is an indicator variable that takes the value one if a news event occurs for a

particular stock, and Short-News Interaction is the product of Short Volume Ratio and the News

Event indicator. Returnt=0 is the return on each stock on the day that short volume and news are

observed. T-statistics are below the parameter estimates in italics and are calculated using

Newest-West (1987) standard errors with 20 lags. *** indicates significance at the 1% level, **

indicates significance at the 5% level, and * indicates significance at the 10% level.


Model

(1) (2) (3) (4) (5)

Intercept 0.0179*** 0.0181*** 0.0179*** 0.0177*** 0.0176***

(4.87) (4.97) (4.92) (5.02) (5.16)


(-2.98) (-2.98) (-2.67) (-2.84) (-2.94)

News Event -0.0012* -0.0002 -0.0002 -0.0002

(-1.48) (-0.26) (-0.20) (-0.17)

Short – News Interaction -0.0049* -0.0055** -0.0056**

(-1.95) (-2.25) (-2.33)

Returnt=0 0.0395** 0.0366**

(2.16) (2.12)

Returnt=-1 0.0236**

(2.34)

Returnt=-2 0.0221

(1.54)


Intercept 0.0061*** 0.0063*** 0.0061*** 0.0060*** 0.0059***

(5.45) (5.27) (5.08) '(4.80) (4.56)


(-3.85) (-3.84) (-3.44) '(-3.59) (-3.66)

News Event -0.0011** 0.0001 0.0002 0.0002

(-1.53) (0.11) '(0.21) (0.23)

Short – News Interaction -0.0063** -0.0069*** -0.0070***

(-2.51) '(-2.84) (-2.94)

Returnt=0 0.0390** 0.0351**

'(2.23) (2.10)

Returnt=-1 0.0258***

(2.70)

Returnt=-2 0.0223

(1.65)

38

Table VI

Cross-Sectional Relation between Returns, Short Sales, and News for Exempt Trades

Table VI contains the results from Fama-MacBeth (1973) type regressions using daily

observations over the period January 3, 2005 through July 6, 2007. The sample only includes

those short sales transactions that were classified as exempt as discussed in Section II.A of the

text. The regressions are done firm by firm, and the dependent variable is the buy and hold

(compound) return over the next 20 trading days. Panel A is calculated using raw returns as the

dependent variable while Panel B uses characteristic adjusted returns as in Daniel, Grinblatt,

Titman, and Wermers (1997), however we omit the book to market factor due to missing

Compustat data for some firms. The Short Volume Ratio is daily short volume / total volume.

News Event is an indicator variable that takes the value one if a news event occurs for a

particular stock, and Short-News Interaction is the product of Short Volume Ratio and the News

Event indicator. Returnt=0 is the return on each stock on the day that short volume and news are

observed. T-statistics are below the parameter estimates in italics and are calculated using

Newest-West (1987) standard errors with 20 lags. *** indicates significance at the 1% level, **

indicates significance at the 5% level, and * indicates significance at the 10% level.


Model

(1) (2) (3) (4) (5)

Intercept 0.0149*** 0.0148*** 0.0147*** 0.0140*** 0.0138***

(4.32) (4.31) (4.32) (4.26) (4.34)

Short Volume Ratio 0.0202** 0.0196*** 0.0203*** 0.0200*** 0.0191***

(5.14) (4.89) (4.77) (4.48) (4.13)

News Event 0.0039* 0.0025 0.0006 -0.0051

(1.90) (0.14) (0.04) (-0.20)

Short – News Interaction 0.2099 0.2189 0.3374

(0.75) (0.83) (0.84)

Returnt=0 -0.0845** -0.1198***

(-2.26) (-3.03)

Returnt=-1 0.1459***

(2.91)

Returnt=-2 -0.1224***

(-3.20)


Intercept 0.0033 0.0032 0.0031 0.0025 0.0022

(1.53) (1.49) (1.47) (1.19) (1.09)

Short Volume Ratio 0.0204*** 0.0200*** 0.0205*** 0.0201*** 0.0191***

(5.75) (5.55) (5.54) (5.15) (4.72)

News Event 0.0030 0.0048 0.0028 -0.0017

(1.62) (0.29) (0.17) (-0.08)

Short – News Interaction 0.158 0.1806 0.2748

(0.64) (0.73) (0.77)

Returnt=0 -0.0815** -0.1198***

(-2.25) (-3.20)

Returnt=-1 0.1602***

(3.58)

Returnt=-2 -0.1407***

(-3.67)

39

Table VII

Equity Returns Following Specific News Events

Table VII examines equity returns following news events according to the model:

where the dependent variable is the compound excess return from day 1 to day 20 following the news event, ret0 is the excess return

on the day of the news event, and Size is measured using the market capitalization for each firm. Regressions are run individually for

each news event and only when a news event occurs. Firm fixed effects are included and the intercept is the average of the fixed

effects. T-statistics are reported below. *** indicates significance at the 1% level, ** indicates significance at the 5% level, and *

indicates significant at the 10% level. Intercept Return (t=0) Short Volume Size

News Events Estimate t-stat Estimate t-stat Estimate t-stat Estimate t-stat

10K 0.0119 0.03 -0.5421 -2.31** 0.0675 0.57 -0.0024 -1.17

8K 0.0223 0.33 -0.1218 -2.20** -0.0079 -0.45 -0.0025 -4.92***

Acquisitions, Mergers, Takeovers 0.0383 0.19 -0.0333 -0.61 -0.0194 -1.26 -0.0067 -5.48***

Analysts' Comments & Ratings 0.0372 0.86 0.0499 1.74* -0.0232 -1.90* -0.0077 -5.89***

Annual Meetings 0.0188 0.96 -0.3073 -1.93* 0.0037 0.14 -0.0012 -2.29**

Antitrust News 0.0118 -0.38 0.3671 3.34*** 0.0051 0.20 -0.0007 -1.42

Bankruptcy-Related Filings 0.0144 -1.55 -0.1910 -1.73* 0.0323 0.86 -0.0016 -2.60***

Bond Ratings & Comments 0.0189 0.37 -0.0117 -0.28 -0.0168 -1.11 -0.0012 -2.50**

Buybacks 0.0186 -0.17 -0.0337 -0.55 -0.0060 -0.29 -0.0009 -2.39**

Contracts, Defense 0.0080 0.33 -0.2923 -1.01 -0.0012 -0.02 -0.0007 -1.08

Contracts, Government (not defense) 0.0247 0.21 -0.0249 -0.11 -0.0300 -0.72 -0.0008 -0.95

Contracts, Nongovernment 0.0323 0.35 -0.1014 -1.24 -0.0229 -1.20 -0.0035 -4.39***

Corporate Governance 0.0416 -1.27 -0.4701 -3.05*** -0.0207 -0.61 -0.0027 -3.64***

Corporate Restructurings 0.0355 0.07 -0.0316 -0.44 -0.0953 -3.24*** -0.0009 -1.65

Divestitures or Asset Sales 0.0325 0.40 0.1591 2.24** -0.0467 -2.29** -0.0022 -3.49***

Dividend News 0.0272 2.04** -0.1339 -4.02*** -0.0047 -0.61 -0.0033 -7.01***

Earnings 0.0412 0.36 0.1145 5.25*** -0.0271 -3.08*** -0.0080 -9.85***

Earnings Projections 0.0445 0.07 0.0594 2.28** -0.0435 -3.76*** -0.0073 -7.48***

Financing Agreements 0.0215 0.58 -0.1209 -1.37 -0.0213 -0.88 -0.0016 -2.81***

High-Yield Issuers 0.0358 1.40 0.0226 0.67 -0.0137 -1.09 -0.0116 -5.68***

Initial Public Offerings 0.0107 -1.41 0.4056 2.91*** 0.0592 2.16** -0.0013 -2.05**

40

Table VII (continued)

Intercept Return (t=0) Short Volume Size

News Events Estimate t-stat Estimate t-stat Estimate t-stat Estimate t-stat

Insider Stock Buys 0.0194 0.13 -0.2852 -3.03*** -0.0012 -0.08 -0.0017 -3.57***

Insider Stock Sells 0.0279 -1.49 -0.2876 -2.18** -0.0208 -0.90 -0.0043 -2.94***

Joint Ventures 0.0318 1.06 0.0844 0.69 -0.0189 -0.75 -0.0028 -3.77***

Labor Issues 0.0372 0.32 0.0647 1.07 -0.0267 -1.25 -0.0034 -4.16***

Lawsuits 0.0417 0.50 -0.0004 0.00 -0.0442 -1.78* -0.0037 -3.78***

Leveraged Buyouts 0.0070 -0.25 -0.2145 -1.26 -0.0018 -0.04 -0.0003 -0.43

Management Issues 0.0382 0.25 0.1289 1.62 -0.0439 -2.06** -0.0040 -5.12***

Market News 0.0416 0.08 -0.0841 -1.68* 0.0054 0.22 -0.0040 -5.36***

Money Market News 0.0375 1.16 -0.6654 -2.71*** -0.0742 -1.22 -0.0006 -1.54

New Products & Services 0.0340 -0.01 -0.1535 -1.31 -0.0819 -4.13*** -0.0021 -2.36**

Personnel Appointments 0.0325 0.22 -0.0694 -1.19 -0.0170 -1.24 -0.0052 -6.20***

Point of View 0.0294 1.55 0.0633 0.43 0.0251 0.65 -0.0040 -2.53**

Product Distribution 0.0260 -1.43 -0.3118 -1.87* -0.0397 -1.11 -0.0013 -2.16**

Research & Development 0.0310 2.34** -0.2311 -1.25 -0.0964 -2.35** -0.0008 -1.31

Spinoffs 0.0315 -0.59 -0.2707 -1.67* -0.0325 -0.40 -0.0011 -1.34

Stock Options 0.0178 0.26 0.0024 0.03 0.0144 0.37 -0.0015 -2.47**

Stock Ownership 0.0318 0.14 -0.0153 -0.35 -0.0225 -3.25*** -0.0042 -8.09***

Stock Splits 0.0234 -0.95 -0.0146 -0.13 -0.0410 -1.74* -0.0006 -1.04

Fisher Stat 160.37***

Fisher P-Value 0.00%

41

Table VIII

Short Volume Portfolio Returns following News Events

Table VIII displays buy and hold portfolio returns for a 12 month period following news events. Each day for each news event, two

portfolios are formed: the first portfolio consists of those firms that had a specific news event and had low short volume as a

percentage of total volume; the second portfolio consists of those that had the news event and had high short volume as a percentage

of total volume. In addition, we form control portfolios using a sample of firms that did not experience the news event but were

similar in terms of bid-ask-spread, institutional ownership, market capitalization, and the number of news events over the previous

month. Difference is the return of the High portfolio less the Low portfolio, and Difference in Difference is the Difference value of the

Control Sample less the Difference value of the Event Sample. *** indicates significance at the 1% level, ** indicates significance at

the 5% level, and * indicates significant at the 10% level. Event Sample: 12 Month Returns Control Sample: 12 Month Returns Difference in

News Events Low High Difference Low High Difference Difference

10K 3.29% 2.23% -1.06% 3.07% 3.99% 0.91% 1.98%

8K 3.56% -1.25% -4.81% 5.52% 3.18% -2.35% 2.46%

Acquisitions, Mergers, Takeovers 4.59% 3.59% -0.99% 4.45% 4.97% 0.52% 1.51%

Analysts' Comments & Ratings 4.61% -0.51% -5.12% 2.33% 4.26% 1.93% 7.05%

Annual Meetings 1.62% -1.25% -2.87% 5.43% 5.57% 0.14% 3.01%

Antitrust News 2.70% 0.81% -1.90% 9.31% 7.51% -1.80% 0.09%

Bankruptcy-Related Filings 5.52% 6.84% 1.32% 8.42% 1.18% -7.24% -8.56%

Bond Ratings & Comments 3.05% 2.17% -0.89% 2.30% 7.29% 4.99% 5.88%

Buybacks 3.77% 0.75% -3.02% 5.32% 3.16% -2.16% 0.86%

Defense Contracts 8.83% 4.37% -4.47% 3.43% 9.59% 6.15% 10.62%

Contracts, Defense 3.25% 5.41% 2.16% 5.68% -2.73% -8.41% -10.57%

Contracts Government (not defense) 3.27% 2.49% -0.78% 3.70% 5.63% 1.92% 2.71%

Corporate Governance 4.33% 2.54% -1.79% 1.21% 9.48% 8.27% 10.06%

Corporate Restructurings 4.19% 4.22% 0.03% 0.34% 9.86% 9.52% 9.49%

Divestitures or Asset Sales 2.15% 2.18% 0.03% 2.27% 6.88% 4.61% 4.58%

Dividend News 4.06% -0.33% -4.39% 3.55% 2.90% -0.65% 3.74%

Earnings 6.47% 1.02% -5.45% 5.84% 2.18% -3.66% 1.79%

Earnings Projections 4.64% 1.33% -3.31% 4.89% 4.02% -0.86% 2.45%

Financing Agreements 2.59% 2.04% -0.54% 4.23% 5.62% 1.40% 1.94%

High-Yield Issuers 3.25% -0.07% -3.32% 6.04% 7.14% 1.10% 4.42%

Initial Public Offerings 5.41% 6.55% 1.14% 5.68% 7.80% 2.11% 0.97%

Insider Stock Buys 0.25% -0.73% -0.98% 5.29% 4.54% -0.76% 0.23%

Insider Stock Sells 1.52% -1.76% -3.29% 4.59% 1.69% -2.90% 0.39%

42

Table VIII (continued)

Event Sample: 12 Month Returns Control Sample: 12 Month Returns Difference in

News Events Low High Difference Low High Difference Difference

Joint Ventures 4.59% 3.62% -0.97% 3.70% 5.59% 1.89% 2.86%

Labor Issues 2.09% 2.91% 0.82% 5.60% 6.45% 0.86% 0.04%

Lawsuits 7.01% 3.50% -3.51% 4.65% 6.34% 1.70% 5.21%

Leveraged Buyouts 1.97% 2.24% 0.27% 5.75% 7.88% 2.14% 1.87%

Management Issues 3.90% -0.63% -4.53% 3.99% 5.30% 1.31% 5.84%

Market News 3.49% 0.95% -2.54% 6.37% 4.44% -1.93% 0.61%

Money Market News 5.73% 7.34% 1.61% 3.68% 19.59% 15.91% 14.29%

New Products & Services 5.22% -0.17% -5.38% 7.95% 4.99% -2.97% 2.42%

Personnel Appointments 5.98% 0.64% -5.33% 3.32% 3.36% 0.03% 5.37%

Point of View 8.11% 0.57% -7.54% 4.09% 4.91% 0.83% 8.36%

Product Distribution 4.03% -2.46% -6.50% 0.67% 6.41% 5.74% 12.24%

Research & Development 3.69% 1.72% -1.98% 7.91% 5.19% -2.71% -0.74%

Spinoffs 1.80% 2.96% 1.15% 4.78% 2.74% -2.04% -3.19%

Stock Options 7.49% 5.31% -2.18% 7.76% 4.43% -3.33% -1.15%

Stock Ownership 2.68% 0.20% -2.48% 5.65% 3.72% -1.94% 0.54%

Stock Splits 4.66% 3.18% -1.48% 2.40% 2.04% -0.35% 1.12%

Mean 2.89%

Median 2.42%

T-statistic 3.73***

Wilcoxon Z Score 3.91***

HOW ARE SHORTS INFORMED? SHORT SELLERS, NEWS, AND … · 2020-07-07 · Kenan-Flagler Business...

Documents

Transcript of HOW ARE SHORTS INFORMED? SHORT SELLERS, NEWS, AND … · 2020-07-07 · Kenan-Flagler Business...